Comment by timshell

This is relatively close to our conclusion from the paper: unless agents are specifically trained for the task and know all the information ahead of time, they're not able to generalize from one cognitive CAPTCHA to another